IP Library › Granted Patent US 9,727,620
Granted Patent B2
US 9,727,620 · App. 14/522,292 · Granted Aug 8, 2017

System and method for item and item set matching

Inventor: Hongwen Kang (San Mateo, CA)
Assignee: Peekabuy, Inc.
G06F17/3053G06F17/30047
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 9,727,620
App. No.
14/522,292
Granted
Aug 8, 2017
Kind
B2
Abstract

A matching system and related methods for matching an input item or input item set with stored items and/or stored item sets, and methods for manufacturing and using same.

Claims (33)

1. A method of item matching implemented on a processor, the method comprising:

receiving a first input item image;

performing image segmentation on the first input item image to generate a segmented input item image;

performing color analysis, texture analysis, and shape analysis on the segmented input item image to generate a plurality of input item-image descriptors;

identifying a plurality of stored item sets stored on a database that includes a second item image, wherein the second item image is similar to the first input item image, wherein said identifying the plurality of stored item sets comprises comparing the plurality of generated input item image descriptors to corresponding item image descriptors associated with the second item image;

ranking the plurality of identified stored item sets;

selecting a portion of the plurality of ranked stored item sets;

replacing the second item image in each of the selected portion of the plurality of ranked stored item sets with the first input item image to generate a plurality of output sets; and

presenting the plurality of generated output sets.

2. The item matching method of claim 1 , further comprising:

associating a plurality of input item-image descriptor categories with the first input item image wherein said identifying the plurality of stored item sets includes comparing the plurality of associated input item image descriptor categories to corresponding item image descriptor categories associated with the plurality of stored item sets.

3. A method of item matching implemented on a processor, the method comprising:

receiving a first input item set including a plurality of input item images;

performing image segmentation on the plurality of input item images to generate a plurality of segmented input item images;

performing color analysis, texture analysis, and shape analysis on the plurality of segmented input item images to generate a plurality of input item image descriptors;

identifying a plurality of stored item sets stored on a database that includes a second item image that is similar to at least one input item image of the plurality of input item images, wherein said identifying the plurality of stored item sets comprises comparing the plurality of generated input item image descriptors to corresponding item image descriptors associated with the second item image;

ranking the plurality of identified stored item sets;

selecting a portion of the plurality of ranked stored item sets;

replacing the second item image in each of the selected portion of the plurality of ranked stored item sets with a corresponding first input item image of the plurality of input item images to generate a plurality of output sets; and

presenting the plurality of output sets.

4. The item matching method of claim 3 , further comprising:

associating a plurality of input item-image descriptor categories with the plurality of input item images, wherein said identifying the plurality of stored item sets includes comparing the plurality of associated input item image descriptor categories to corresponding item image descriptor categories associated with the plurality of stored item sets.

5. A method of item matching implemented on a processor, the method comprising:

receiving a first input item set including a plurality of input item images;

performing image segmentation on the plurality of input item images to generate a plurality of segmented input item images;

performing color analysis, texture analysis, and shape analysis on the plurality of segmented input item images to generate a plurality of input item image descriptors;

identifying, for each input item image, at least one similar stored item image within a plurality of stored item images on a database, wherein said identifying comprises comparing the plurality of generated input item image descriptors to corresponding item image descriptors associated with the plurality of stored item images;

replacing the plurality of input item images with a respective identified similar stored item image in various combinations to generate a plurality of generated item sets;

ranking the plurality of generated item sets;

selecting a portion of the plurality of ranked generated item sets to generate output sets; and

presenting the output sets.

6. The item matching method of claim 5 , further comprising:

associating a plurality of input item-image descriptor categories with the plurality of input item images, wherein said identifying, for each input item image, at least one similar stored item image includes comparing the plurality of associated input item-image descriptor categories to corresponding stored item-image descriptor categories associated with the plurality of stored item image sets.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 25, 2021
From: PEEKABUY, INC.
To: HANGZHOU HUICHUAN TECHNOLOGY CO., LTD
Reel/Frame 055013/0227 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 23, 2014
From: KANG, HONGWEN
To: PEEKABUY, INC
Reel/Frame 034023/0031 →
Continuity (2)
Provisional Application 61897056 · Oct 29, 2013
Related Publication 20150120759A1 · Apr 30, 2015